Issue with background type being changed in unknown after calibration

[Mike Jercinovic]

Okay, here is a new one.  Running monazite, as usual, but my Pb values looked just a bit low on our reference standard.  So, let's recalibrate Pb and go from there.  This is a two spectrometer integration for PbMa, and the analysis is done with multipoint background acquisition, TDI, Nth pt., and on-peak time fraction.  So off I go to our Pb standard and calibrate the way we always do, turn off TDI and on peak time fraction, and two point "regular" backgrounds.  Nice re-calibration, slightly lower intensities at the same peak positions on the two spectrometers so should increase the Pb in the unknowns just slightly, awesome! So now I want to see what effect this has on the previous run of the reference monazite.  I click analyze for this "unknown", and much to my surprise, it gives a Pb value that is now WAY wrong, much too low by 100ppm or more.  What is going on?  So I go through everything.  Looks good, but the net intensity values are all way low, but just for Pb.  So now go to look at the multipt. arrays.  Somehow now these have been changed to linear for Pb in the unknown.  So I change them back to exponential with the correct amount of points for iteration.  Try it again.  Still wrong, no change!?  So now have a hunch this is screwed up elsewhere so I go into elements/cations for this unknown in the analyze window.  Sure enough, the background type is now switched to two point linear here for my two Pb acquisitions!  So now I change this back to multipoints for both, with the appropriate number of points for iteration and try again.  Now everything is good, and my numbers look great.  This is a typical procedure for us, so this sort of behavior is quite new.  Just calibrating a standard has managed to screw up my previous unknown (but only this one - the other unknowns are fine!) This is version 12.3.7.

[John Donovan]

Okay, here is a new one.  Running monazite, as usual, but my Pb values looked just a bit low on our reference standard.  So, let's recalibrate Pb and go from there.  This is a two spectrometer integration for PbMa, and the analysis is done with multipoint background acquisition, TDI, Nth pt., and on-peak time fraction.  So off I go to our Pb standard and calibrate the way we always do, turn off TDI and on peak time fraction, and two point "regular" backgrounds.  Nice re-calibration, slightly lower intensities at the same peak positions on the two spectrometers so should increase the Pb in the unknowns just slightly, awesome! So now I want to see what effect this has on the previous run of the reference monazite.  I click analyze for this "unknown", and much to my surprise, it gives a Pb value that is now WAY wrong, much too low by 100ppm or more.  What is going on?  So I go through everything.  Looks good, but the net intensity values are all way low, but just for Pb.  So now go to look at the multipt. arrays.  Somehow now these have been changed to linear for Pb in the unknown.  So I change them back to exponential with the correct amount of points for iteration.  Try it again.  Still wrong, no change!?  So now have a hunch this is screwed up elsewhere so I go into elements/cations for this unknown in the analyze window.  Sure enough, the background type is now switched to two point linear here for my two Pb acquisitions!  So now I change this back to multipoints for both, with the appropriate number of points for iteration and try again.  Now everything is good, and my numbers look great.  This is a typical procedure for us, so this sort of behavior is quite new.  Just calibrating a standard has managed to screw up my previous unknown (but only this one - the other unknowns are fine!) This is version 12.3.7.

Hi Mike,
Could you please describe a more complicated combination of methods and sequence of events!   

Since we haven't changed any of that code in quite some time (~1 year or more), I suspect that this is just a particular unique sequence of events that caused your bgd model to get changed.  Frankly I have no idea how acquiring a new standard would change the background on your unknown, unless you somehow had selected both stds and unkns at some point.

Anyway, I think we should take this chat off-line and have you send me your MDB file and then we can see if we can reproduce the issue.  Which I suspect we won't be able to!   
john